Go to Lecture One or Twoor Three or Four or Five or Six or Eight

Go to Cyber Semiotic Institute home page
Go to Course Outline

Human Communication as a Primate Heritage

Instructor: Anne Zeller

Lecture Seven: Development of Language Use

When did human beings begin to talk? This question is probably one of the oldest that we ask ourselves after "Where did we come from". The two are probably closely intertwined. One of the things that makes us human is our unique ability to utilize a complex verbal system for the storage and transmission of abstract information. I say unique, yet I know that the argument about this uniqueness rests on a controversy about what words really are and how complex a word use system has to be before it can be called language. Is comprehension of single spoken words enough? In this case most dogs could be said to have some language use. Are both reception and production required? In this case injured individuals who cannot produce words anymore can be said to have lost their language. If reception and production at a simple level of gestural indication are enough, can apes be considered to have language use if they have been trained in a human signing or symboling system?

The focus of this lecture will be on what evidence we have in the evolutionary past for the physical and cultural indicators of language use. We will included some discussion of the parameters of ape sign and symbol use as part of the discussion on what mental and brain size attributes are necessary to reach certain levels of verbal interchange.

Let us begin with the size of the brain. The common ancestor of apes and hominids has not yet been found, but the earliest hominids (Australopithecus) forms have a cranial capacity very similar to that of modern chimpanzees. This is around 350-400 cubic centimetres (or c.c.). This is enough to run a 30 to 40 kilo mammalian body and still have some left over. Big dogs can also weigh 30 to 40 kilos, but have much smaller cranial capacities. The left over brain capacity in primates is available for learning, memory, and complexities of behaviour. I should mention that there is some variability in size between the sexes, (sexual dimorphism) in both brain and body size, and in fact this was much more evident in early hominids than it is now. However the overall brain: body ratio is fairly similar between modern males and females (Dixson, 1981). The range in ape cranial capacity for gorillas is: females 340-595 c.c. and males 470-752 c.c. with means of 458 c.c. and 535 c.c. respectively. I chose gorillas here as an example because their level of sexual dimorphism is much more similar to early Australopithecus afarensis, (2:1) than is the chimpanzee. Modern chimpanzees and modern humans have about the same level of sexual dimorphism (a difference of about 12%). However, modern chimpanzees have cranial capacities for females of 278-455 c.c. with a mean of 355 c.c. and for males of 322-500c.c. with a mean of 396 c.c., with the same range of body sizes as was found in early Australopithecus (Dixson, 1981). The early sexually dimorphic Australopithecus afarensis had cranial capacities ranging from 350 (female) to 500 c.c. (male), which, as you can see puts them in the range of modern chimpanzees. Thus, in terms of overall size and brain to body ratio, they should have similar quantities of brain available to devote to communication.

The problem is that overall size is probably not as important as complexity. One of the major features of human brains that is associated with language use is brain lateralization. In other words, the two halves of the brain do not mirror each other completely. In particular, language based brain functions such as motor control and meaning/association for words are focused on the left side of the brain and contribute to brain asymmetry. The asymmetry can be recognized in size differences of the two halves. Early work by Tobias (1971) and others suggest that the beginning of brain asymmetry in early Australopithecus exists. There are current studies on chimpanzees that suggest that there are certain very specific lateralized brain functions having to do with language which are present in chimpanzee brains. In particular is the 'planum temporale' which is a specialized area in the auditory association cortex which receives sounds, and attaches meaning to them. In most of the chimp brains tested, this area was larger on the left side hemisphere than in the right one, which is also the human pattern (Begley, 1999). This implies a similarity of pattern between ape and human which may underly specialized sound recognition. Even though chimpanzees have a vocal system of calls it requires much less specialized memory and association to receive and produce these calls than it does to manipulate the intricacies of human language. However, chimpanzees can learn to use gesture based sign systems such as American Sign Language or pictoral based symbol system such a Yerkish. Up to 400 to 500 patterns can be recognized by trained chimpanzees, and gorillas who can both comprehend and produce them, although they do not have the anatomical structure and motor control over their lips and tongues to speak. Their level of language comprehension, however, has been determined to be quite high. Kanzi, a bonobo (which is a species of chimpanzee) who has been in a language learning environment for nearly 20 years has a very high level of comprehension of spoken English. He was tested on 660 novel English sentences and responded to the correctly on 72% of the trials, which was an equivalent response rate to that of a 2 year old human child tested on the same problems. It seems as though Kanzi can discriminate the meaning of many English words and respond appropriately to them even when they are embeded in a fairly complex syntax. (Savage-Rumbaugh et at. 1998). His major problem seems to be in keeping several unrelated items in this short term memory at the same time. He had trouble when asked to pick up several different items and show them to the researcher at the same time. He also had trouble when words that sound the same were used in different ways. An example would be "Get a can of coke", "Can you touch your nose?" and "Put this in the trash can." However, when the context was very clear he was able to distinguish words such as "dog", and "hot" and "hotdog" quite well. (Savage-Rumbaugh et al. 1998:70). The little girl age 2 The little girl age 2 1/2 who was concurrently being tested also had some difficulty in responding correctly when asked to bring one or more objects. She did not seem to grasp numerical factors well. She was also confused by words which sounded very similar to each other. In Kanzi's case the structural aspects of sentence composition actually made it easier for Kanzi to understand and to remember what he was supposed to be doing. This suggests that in addition to having specialized abilities to hear and comprehend the differences in sound that make up words, he has the ability to sort out the associations between them. This is an aspect of expansion of association areas and increased inter-cortical connections.

The changes in the brain structure which allow these associations to occur have only begun in apes, but show an ever increasing presence in the brains of developing hominids. Obviously we do not have real fossilized brains of early forms, but we do have indications of their size and the complexity of lobes from the patterns that are left on the inside of skulls from casts called endocasts. These patterns suggest that cortical association areas on early hominids were rapidly developing. This is particularly important in the posterior parietal region which is the location of a major source of speech symbols and concepts, and includes Wernike's area. This part of the brain is highly lateralized in function in humans, with speech functions being in the left hemisphere and complex spatial and body movements being controlled on the right side. The development of the parietal region into a larger area really did not begin until hominids had reached the level of Homo erectus. When you look at an Australopithecus skull, the area available for the parietal region is really very small. As Homo habilis (beginning about 2.4 million years ago) develops, the brain does begin to expand, but the shape is still fairly similar to the Australopithecus. What this could mean is that they could probably hear and distinguish and organize the sounds and gestures of their world as well (or better) than chimpanzees can, but quite possibly could not produce the complex verbal signals that make up language. By the time of Homo erectus -- about 1.8 million years ago the cranium had expanded to 800-900 cc. and much of that expansion was in the parietal and occipital regions of the brain. Not only was the brain getting bigger, but its asymmetries were becoming more developed. In correlation with this the 'corpus callosum' or part of the brain that connects the two hemispheres is much larger in humans than in other animals and provides instantaneous communication complete integration of the various functional areas, such that hemisphere asymmetries are smoothly coordinated (Campbell, 1985).

The area of cortex that most clearly differentiates ape from human brains in the frontal lobes. It is here, on the left side that "Broca's area' is found. This is an area, specialized to the left hemisphere, which is very important in organizing and permitting speech production. There is some controversy over whether Australopithecus has a Broca's area, but it certainly has a smaller frontal lobe which is much less convoluted. In other words, the frontal lobe of Australopithecus africanus brains are much more ape-like than human-like in both size and sulcus pattern (Falk, 1992). The specialized area for speech production is not clearly indicted in the brain endocasts, leaving it open to argument if one was there.

This pattern of the presence of Wernike's area for sound comprehension in the temporal area, and the possible absence of a specialized area for speech production, in apes and Australopithecus correlates quite well with the anatomical attributes of the mouth region required for speech. Although soft tissue such as lips and tongues do not fossilize, there is some evidence in the hard structures we do have. In particular, the size and location of the tongue are very important in governing the articulate nature of the sounds that we make. In humans the tongue is a short fat organ which is remarkably mobile. It can move around the space between the mouth floor and palate very precisely controlling the shape and size of the resonating chamber and the flow of air through it. This is what allows the sounds to be clear and unambiguous in humans. The difference between humans and chimpanzees is in three areas. The chimps have longer thinner tongues which are not as effective at stopping the air flow and managing the shape of the oral cavity in order to make precise sounds. The second difference is that the chimpanzee larynx is placed higher in the throat which allows them to breath and swallow at the same time. Human infants who need to nurse also have their larnyx in a high position, and it does not move down the airway until they are over a year old and nursing much less frequently. The corollary of having a high larynx is that the nasal passage cannot be blocked off from the mouth cavity. This means that air being expelled from the lungs for purposes of vocalizing comes out both the nose and mouth, thus making sounds which are much 'fuzzier'. This was one of the factors which the Hayes had to deal with when trying to teach the young chimpanzee "Vicki" to speak. They had to put their fingers over her nostrils in order to focus the sound flow through her mouth, in an effort to get her to enunciate words. This was not a very successful experiment in terms of results (Hayes and Nissen, 1971). The third anatomical difference is that chimpanzees have an air sac in their throats and can direct air though the vocal cords from this as well as from their lungs. The this can influence the vibration patterns of the vocal cords to such an extent that clear noises do not ensue.

Most of these features are soft tissue and do not preserve in the fossil australopithecus being studied. However, the one aspect that is preserved is the depth (or height) of the palate. In forms with a high arched palate, one can assume the short fat type of tongue which correlates with language production. In the gracile types of australopithecus the palate is rather shallow, which indicates that the tongue was probably thin. The information about larynx position and vocal sac cannot be ascertained from fossils, but palate depth is an important indicator of controlled sound production. In the transition to Homo we begin to see some structural changes towards a possible specialized sound production ability including increased height of the palate. Another major change is in the frontal lobe region of the brain endocasts. In Homo habilis -- the convolutions in the frontal lobe is much more like a human than it is like an apes. Tobias long ago had suggested (1971) that H. habilis might have a Brocas area and Falk (1992) feels that by specialized measuring techniques she has confirmed its presence. This goes along with the beginning of brain expansion up to between 600 and 750 cc., the increase of the parietal areas and the increasing height of the palate. These structural features of both brain and body may correlate with a transition to somewhat more articulate calls or sounds. This may not yet be speech, but it may serve as its foundation.

In addition to brain size, complexity, lateralization and specialized sound comprehension and production areas and vocal tract, characteristics another important correlate to speech is the development of complex neuromuscular coordination for the fine level of motor control required for speech production. We do not have much preserved evidence for motor control in the body. One place to look would be in the size of holes for nerves to pass through. The major one of these is the spinal cord. In the Homo erectus fossil boy WT. 15000. The hole in the vertebrae for the spinal cord is relatively small. One person has suggested that this implies lack of nerve control required to coordinate breathing with talking. This would suggest that Homo erectus had rather less ability to produce controlled vocalizations than later forms. Another nerve canal (the hypoglossal) under investigation, runs through the base of the skull and its is argued that a larger diameter implies more nerve control over the muscles that govern speech. However, the evidence on size variability on current forms of ape and human is still not entirely collected, so the proposed correlation has yet to be established.

One thing that we do have, however, is the beginning of stone tools. The current evidence suggest that about 2.4 million years ago early hominids made the transition from using sticks, vines and unmodified stones as tools, in the pattern of modern chimpanzees, to actually modifying stone to improve its function as a tool. Chimpanzees and orangutans have been taught to do this, but they have not come up with the idea on their own in the wild. They often just smash the rocks by throwing them on a hard substrate and get a sharp edge which can be used for cutting. Situations in which flakes are purposefully struck from a stone with another stone require planning, hand-eye coordination and an understanding of how a problem can be solved. The pattern of flake production on Oldowan tools has been closely examined which has resulted in ascertaining some interesting differences from stone tools made by chimpanzees. In the first place stone was being transported for use as a tool over quite long distances (up to 10 km) which is well beyond the distance that chimpanzees will currently carry a tool (Toth and Schick, 1993). Second, some principles of stone flaking, such as the need to strike the core where the stone is already at an acute angle in order to detach a useable flake seem to be well understood. In addition, the clockwise direction of the rotation of the core when striking successive flakes from it, strongly suggest that it was held in the left hand and struck with a stone in the right hand (Toth and Schick, 1993). Hand lateralization is not common in chimpanzee feeding or locomotion but is strongly evident in most tool use and skilled food processing patterns. (McGrew and Marchant, 1996). In other words, although in general there is mixed evidence for handedness in chimpanzees, there is much more evidence for lateralization, in the skilled precise hand movements required in termite fishing or nut cracking with stone hammers.

The reason why hand laterality and particularly right handedness is important is that in the differentiation of function seen in a lateralized brain, the primary motor cortex is involved. The extra development of the frontal lobe which houses the primary motor cortex in a modern human is spatially associated with the specializations for speech which are also located there. The premotor cortex which is directly in front of the primary motor cortex, "seems to be involved with accessing the memory of complex movements that have been learned" (Falk 1997:62). These are complex learned sequences of events, such as tooth brushing or piano playing, or other strings of learned behaviour, such as making stone tools. The prefronal cortex is also responsible for determining the position of objects in space and integrating behaviour guided by representational memory. It uses this stored representational knowledge to guide motor responses in a planned sequence of actions because it is also responsible for sustained attention in human behaviour (Campbell, 1985). Chimpanzees have difficulty in keeping a number of nonmaterial factors in mind at one time (remember Kanzi, not being able to remember the items in a list) (Savage Rumbaugh et al., 1998) and this would be a major hinderance to the planning necessary to assemble a core, a hammer stone and the sequence of motor events necessary to turn a rock into a tool. While chimpanzees are much stronger than people, their ability to concentrate on a long series of fine motor movements is not nearly as high as a human's. Thus, the increased frontal cortical region, left hemisphere specialization controlling fine motor movements and evident use of the right hand to strike the core precisely in order to to flake it, are all part of a package. Hand-eye coordination must be quite precise in achieving this without injury to oneself. (If you doubt me on this just try for yourself. I personally had a cast on my left thumb for weeks after trying to demonstrate how to make an Oldowan stone tool!) A high level of sensory-motor feedback and precise motor control are required to keep long stings of precise rythmic motor movements in mind and correctly respond to the changes in the tool you are working on. These aspects of motor control and mental concentration parallel the type of memory, co-ordination and motor control required to energize the breathing apparatus, vocal cords, tongue and lips into the sequence of events required to enuciate a word. The level of motor control required to make stone tools may be either a precursor to, or a parallel with the motor control sequence which allows articulate speech.

Very few people would attribute coherent verbal speech to Homo habilis. Their brains are small, their palates only beginning to arch, and the level of complexity seen in the motor patterns required to make Oldowan tools is not high. However, Homo habilis was very much a transitional form of hominid. Within a few hundred thousand years of its emergence a newer type of Homo began to develop. This was Homo erectus, mentioned above, with his 800-900 c.c. brain, which is developing in the parietal area in particular. The parietal region houses a major association area and in modern humans is about twice as large as would be estimated for a primate of our body size (Campbell, 1985). This region takes in sensory input from the skin, the auditory region, and visual input and coordinates the information, not only within each area but between them. In many ways it seems that the specifically human nature of the brain lies in the refined level of interconnection of its parts. This allows the human to process incoming information at a variety of levels up to the abstract, and to respond to it in a variety of ways. Laterality of response systems is a major attribute in adapting to a wide variety of features in the habitat. One major feature of habitat change that is occurring as Homo erectus is developing is the changeover in tool types from the simple, all purpose Oldowan style stone tools, to the more specialized, complex and stylized tools of the Acheulian tradition, which allow a much wider response to habitat manipulation.

Acheulian tools become evident in the archaeological record about 1.6 million years ago, which is close to the time at which Homo erectus developed. They are made in a symmetrical fashion, with a smooth cross section, balanced form and straight edges. Even modern people can take months to learn how to choose the correct material, rough out a suitable core or large flake, construct striking edges of a suitable acute angle, and finish the edge flaking with soft percussion, using wood or bone hammers. The forms are very normative in terms of size, shape and finishing techniques, although these norms develop over time. This tool assemblage lasted for over 1 million years and spread from Africa to Europe and Western Asia. The details of construction, the initial choices made of material, the roughing out for a particular tool, -- cleaver, chopper, scraper, pick etc. and the finishing details could all be learned by observation with little vocal input. However the frequency of production, the spread of the pattern and the anatomical potential for more extensive vocal production suggest that some level of vocal communication may well have been involved. As well, the level of motor control involved would provide a pattern for the motor control involved in producing hominid speech.

It is the level of control, timing and precision of Acheulian tool production that argues most strongly for these factors being widespread in the lives of Homo erectus. They are the first hominids who moved out of Africa and erectus fossils have been found in the Near East, Georgia and Java dating to about 1.6 million years ago. Toth and Schick (1993) argue that the lack of Acheulian tools in the far east is associated with the loss of the technology due to inadequate raw

materials. They went through regions that did not have appropriate materials and lost the ability to make them, since it was not stored in a cultural communicative form which allowed retrieval of information from the past. (Toth and Schick, 1993) I would prefer to argue that the far eastern Homo erectus had left Africa before the Acheulian tool making pattern was invented since they are present in Asia by at least 1.6 million which is the accepted early date for this tool development in Africa. The relevance of this is the argument that such a complex learning and memory based procedure had to be preserved and taught via complex communicative system, even if it was only partially linguistic. It was not so simple that it could be reinvented over and over again in various places all the same way. You have Acheulian tools spread widely, but their origins localized in time and space. The transmission of information to make them was learned, and I would argue was coded in a language system. At this point Homo erectus had about 2/3 of modern cranial capacity, a lateralized asymmetrical brain, a high arched palate, a group living situation, and a fairly complex stone tool producing culture when compared with any animal currently alive. When these accomplishments are compared with those of great apes who have social groups, simple tools, asymmetrical brains and complex sound perception abilities, the major differences seem to be aspects of brain size and structure, and differences in memory, concentration, abstraction and precision sound production capabilities.

I realize that many other scholars would like to argue that human language is a preserve of Homo sapiens (whether or not you include Neandertals). I have argued that the beginnings of human style language focusing on labelling, action, and simple structural (syntactic) associations between words and meaning were present in Homo erectus. The level of gestural skill required to sign a vocabulary of up to 400-500 words is within the range of common signing chimpanzees with brains half the size of and much less complex than Homo erectus. A combination of gesture and vocalization is an undoubted basis for early hominid communication systems. As the brain became larger, more convoluted, more lateralized and more specialized, motor control over the whole body increased. As sound production became clearer, gesture could become more refined and more culturally constrained. It is clear from the tenor of these lectures until now that humans have not abandoned their nonverbal gesture and communication system. Rather their vocal-auditory system has become specialized to convey the specialized type of information humans find essential to the world they have constructed. This type of information is often abstract, time based, conjectural, planning information used to record the past and organize the future. It is about mentalistic functions such as consciousness, love, death and the stock market. It is a life system far outside the one that the other animals on this planet inhabit, and we need our own specialized communication system to inhabit the world we have made. This system is as culturally controlled as any of our current subsistance and political systems, but like them it had its roots in the life patterns of our animal ancestors foraging and group living on the plains of the past. Modern language is a result of modern culture, but language as a communication system of verbal signs set in a simple but informative syntax has been with us, in my view, for over a million years.


Begley, S. (1999) "Aping Language." In. Annual Editions. Physical Anthropology 99/00 ed. E. Angeloni. McGraw-Hill. Conn. pp 64-66.

Campbell, B. (1985) 3rd ed. Human Evolution, Aldine de Gruyter. New York.

Dixson, A.F. (1981) The Natural History of the Gorilla, Columbia University Press. New York.

Falk, D. (1992) Braindance. Henry Holt & Co. New York.

Hayes, K.J. and Nissen. C.H. (1971). "Higher mental functions of a home raised chimpanzee." In: Behavior of Non-Human Primates. Vol. 4. ed. A.M. Schrier and F. Stollnitz Academic Press. New York. pp. 60-115.

McGrew, W.C. & Marchant, L.F. (1996). "On which side of the apes? Ethological Study of Laterality of Hand Use." In: Great Ape Societies. eds. W.C. McGrew, L.F. Marchart & T. Nishido. Cambridge University Press. Cambridge, Mass. pp 255-272.

Savage-Rumbaugh, S. and Shanker, S.G. and Taylor, T.J. (1998). Apes, Language, and the Human Mind. Oxford University Press. New York.

Tobias, P. (1971). The brain in hominid evolution. Columbia University Press. New York.

Toth, N. and Schick, K. (1992). "Early stone industries and inferences regarding language and cognition." In: Tools, Languge and Cognition in Human Evolution. eds. K.R. Gibson and T. Ingold. Cambridge. University Press. Cambridge pp. 346-362.

copyright 1999 Anne Zeller
Send comments or question to Anne Zeller: azeller@artspas.watstar.uwaterloo.ca
Go to Lecture One or Twoor Three or Four or Five or Six or Eight
Go to Cyber Semiotic Institute home page
Go to Course Outline